Databases, Chapter 10

A property that requires all parts of a transaction to be treated as a single, logical unit of work in which all operations must be completed (committed) to produce a consistent database.

Not being able to be divided into smaller units. In transaction management, the transaction property that requires all parts of a transaction to be treated as a single, indivisible, logical unit of work. All parts of a transaction must be completed or the entire transaction is aborted.

A lock that has only two states: locked (1) and unlocked (0). If a data item is locked by a transaction, no other transaction can use that data item.

Temporary storage area in primary memory used to speed up disk operations.

In transaction management, an operation in which the database management system writes all of its updated buffers to disk.

A DBMS feature that coordinates the simultaneous execution of transactions in a multiprocessing database system while preserving data integrity.

A database condition in which all data integrity constraints are satisfied. To ensure consistency of a database, every transaction must begin with the database in a known consistent state. If not, the transaction will yield an inconsistent database that violates its integrity and business rules.

A database state in which all data integrity constraints are satisfied.

The equivalent of a single SQL statement in an application program or a transaction.

A type of lock that restricts database access to the owner of the lock and allows only one user at a time to access the database. This lock works for batch processes but is unsuitable for online multiuser DBMSs.

A condition in which two or more transactions wait indefinitely for the other to release the lock on a previously locked data item. Also called deadly embrace.

In transaction management, a condition in which transaction operations do not immediately update a physical database. Also called deferred write technique.

In transaction management, when a transaction reads data that is not yet committed.

In permanent storage, the equivalent of a disk block, which can be described as a directly addressable section of a disk. A diskpage has a fixed size, such as 4K, 8K, or 16K.

The transaction property that ensures that once transaction changes are done and committed, they cannot be undone or lost, even in the event of a system failure.

An exclusive lock is issued when a transaction requests permission to update a data item and no locks are held on that data item by any other transaction. An exclusive lock does not allow other transactions to access the database.

A lock that allows concurrent transactions to access the same row as long as they require the use of different fields (attributes) within that row. This type of lock yields the most flexible multiuser data access but requires a high level of computer overhead.

In concurrency control, a process that ensures a database is immediately updated by operations during the transaction’s execution, even before the transaction reaches its commit point.

A concurrency control problem that arises when a transaction-calculating summary (aggregate) functions over a set of data while other transactions are updating the data, yielding erroneous results.

A database transaction property in which a data item used by one transaction is not available to other transactions until the first one ends.

A device that guarantees unique use of a data item in a particular transaction operation. A transaction requires a lock prior to data access; the lock is released after the operation’s execution to enable other transactions to lock the data item for their own use.

The level of lock use. Locking can take place at the following levels: database, table, page, row, and field (attribute).

A DBMS component that is responsible for assigning and releasing locks.

A concurrency control problem in which a data update is lost during the concurrent execution of transactions.

A quality that ensures that time stamp values always increase. (The time stamping approach to scheduling concurrent transactions assigns a global, unique time stamp to each transaction. The time stamp value produces an explicit order in which transactions are submitted to the DBMS.)

A condition in which only one transaction at a time can own an exclusive lock on the same object.

In transaction management, a concurrency control technique based on the assumption that most database operations do not conflict.

In permanent storage, the equivalent of a disk block, which can be described as a directly addressable section of a disk. A diskpage has a fixed size, such as 4K, 8K, or 16K.

In this type of lock, the database management system locks an entire diskpage, or section of a disk. A diskpage can contain data for one or more rows and from one or more tables.

The use of locks based on the assumption that conflict between transactions is likely.

In transaction management, when a transaction executes a query at time t1, then runs the same query at time t2, yielding additional rows that satisfy the query.

An ANSI SQL transaction isolation level that allows transactions to read only committed data. This is the default mode of operations for most databases.

Multiple copies of the transaction log kept by database management systems to ensure that the physical failure of a disk will not impair the DBMS’s ability to recover data.

An ANSI SQL transaction isolation level that uses shared locks to ensure that other transactions do not update a row after the original query updates it. However, phantom reads are allowed.

A less restrictive database lock in which the DBMS allows concurrent transactions to access different rows of the same table, even when the rows are on the same page.

The DBMS component that establishes the order in which concurrent transaction operations are executed. The scheduler interleaves the execution of database operations in a specific sequence to ensure serializability.

A property in which the selected order of concurrent transaction operations creates the same final database state that would have been produced if the transactions had been executed in a serial fashion.

An ANSI SQL transaction isolation level that does not allow dirty reads, nonrepeatable reads, or phantom reads; the most restrictive level defined by the ANSI SQL standard.

In transaction management, a schedule of operations in which the interleaved execution of the transactions yields the same result as if they were executed in serial order.

A lock that is issued when a transaction requests permission to read data from a database and no exclusive locks are held on the data by another transaction. A shared lock allows other read-only transactions to access the database.

A locking scheme that allows only one transaction at a time to access a table. A table-level lock locks an entire table, preventing access to any row by transaction T2 while transaction T1 is using the table.

In transaction management, a technique used in scheduling concurrent transactions that assigns a global unique time stamp to each transaction.

A feature used by the DBMS to keep track of all transaction operations that update the database. The information stored in this log is used by the DBMS for recovery purposes.

A set of rules that governs how transactions acquire and relinquish locks. Two-phase locking guarantees serializability, but it does not prevent deadlocks. The two-phase locking protocol is divided into two phases: (1) A growing phase occurs when the transaction acquires the locks it needs without unlocking any existing data locks. After all locks have been acquired, the transaction is in its locked point. (2) A shrinking phase occurs when the transaction releases all locks and cannot obtain a new lock.

A concurrency control problem in which a transaction accesses uncommitted data from another transaction.

In concurrency control, a property of time stamping that ensures no equal time stamp values can exist.

A concurrency control scheme in which an older transaction must wait for the younger transaction to complete and release the locks before requesting the locks itself. Otherwise, the newer transaction dies and is rescheduled.

A concurrency control scheme in which an older transaction can request the lock, preempt the younger transaction, and reschedule it. Otherwise, the newer transaction waits until the older transaction finishes.

In concurrency control, a process that ensures transaction logs are written to permanent storage before any database data is actually updated. Also called a write-ahead protocol.

In concurrency control, a process that ensures a database is immediately updated by operations during the transaction’s execution, even before the transaction reaches its commit point. Also called immediate update.

In transaction management, when a transaction reads a given row at time t1, then reads the same row at time t2, yielding different results because the original row may have been updated or deleted.

An ANSI SQL transaction isolation level that allows transactions to read uncommitted data from other transactions, and that allows nonrepeatable reads and phantom reads. The least restrictive level defined by ANSI SQL.